|
|
Post by UnklMickey on Oct 6, 2006 16:29:37 GMT -5
a buffer built into a cable? unk sez:  |
|
|
|
Post by ChrisK on Oct 6, 2006 18:22:59 GMT -5
|
|
|
|
Post by UnklMickey on Oct 6, 2006 21:03:39 GMT -5
i doubt i'd ever be in the market for 2000 of these puppies. but, anything that comes in an "ammo pack", really gets your attention. |
|
|
|
Post by JohnH on Oct 6, 2006 21:35:49 GMT -5
Im not too sure what we are talking about in these last two posts. But if it is related to inconsistency in the specs of MPF102, then I'm not worried about it for this circuit, particularly as a one off build, because I believe it to be reasonably tolerant of variations in the JFET, within at least most of the spec range.
John
|
|
|
|
Post by UnklMickey on Oct 6, 2006 22:59:18 GMT -5
...I believe it to be reasonably tolerant of variations in the JFET, within at least most of the spec range. John that stands to reason, for a depletion mode drain follower, with the gate referenced to ground, and such a small input signal. i wasn't casting any shadows. i was just musing over the "ammo box" terminlogy in the packaging. |
|
Channelman
Meter Reader 1st Class
Posts: 62
Likes: 0
|
Post by Channelman on Oct 7, 2006 4:54:34 GMT -5
I don't think that the MPF102 is the ideal choice in a buffer running from a 9v supply.
The Vgs required to bias it down to 200uA could be as high as 7.5v.....according to the spec.
We could have a serious 'headroom' problem but (as JohnH implies...they probably won't be
at the top end of the spec). But any with spec over 4v will have a problem.
My vote is for the J201...specced as '0.3 to 1.5volts', especially if we are after some voltage gain.
EDIT
Sorry, just realised that the MPF102 is for the 'cable buffer' only which has different bias arrangement....
It should work OK.
Does anybody know of an 'enhancement mode' N channel fet which is readily obtainable?
These usually 'turn on' about 1 to 1.5 volts and would appear to be useful in this and similar applications.
CM
|
|
|
|
Post by JohnH on Oct 7, 2006 7:44:40 GMT -5
Thanks CM - the MPF102, or something similar with a higher Vgs, is the only way I could think of getting the JFET biased adequately with just a resistor to ground. Anything that also needs a resistor to positive, then also needs an input cap, and the jack-plug concept then gets too hard to fit in. The design specs for these JFETs are very loose (thanks for posting them Chris). The article on the Fetzer valve has some tests on gate/source bias voltages for quite a good number of examples of typical JFETs: www.runoffgroove.com/fetzervalve.htmlFor the MPF102, it found a range of 1.76V to 3.11V out of 24 examples, which is a very suitable range for the buffer cable. During the day I have been adding more description and info to the buffer cable part, based on suggestions from Unk. cheers John |
|
Channelman
Meter Reader 1st Class
Posts: 62
Likes: 0
|
Post by Channelman on Oct 7, 2006 8:14:19 GMT -5
Thanks John. Just ignore the rambliings in my last post... I edited it when I realised what it was for. I hadn't gone back and seen your 'cable buffer'. All the range will be OK. My requested for an 'enhancement mode N Channel FET still stands though. I think it would be useful in battery circuits because the gate is more positive than the source and we wouldn't need to 'waste' supply voltage 'dropping' the required negative Vgs across a source resitor. We should finish up with more 'headroom' depending on the design. I know of a FET (ZVN3306A) but not a readily available one. www.zetex.com/3.0/pdf/ZVN3306A.pdfCM |
|
|
|
Post by JohnH on Oct 7, 2006 8:46:20 GMT -5
CM - they sound interesting, I wonder what a typical circuit for an enhancement mode FET would be like?, and how they would behave when pushed?
My wish list would include a FET withe Vgs of about -1.5V, with a tight consistency to save having to test each one
John
|
|
Channelman
Meter Reader 1st Class
Posts: 62
Likes: 0
|
Post by Channelman on Oct 7, 2006 8:51:34 GMT -5
Thats just it John. This example is tighter. Vgs 0.8v to 2.4v but positive instead of negative and usually they are 1 to 1.5 volts. My reasoning was that with the source down at ground, the device could be biassed such that the drain was at mid-rail, giving us a potential output 'swing' of 9 volts peak to peak. None would be wasted 'generating' a 'relative' negative supply for the gate by elevating the source. The positive bias for the gate would come from the positive rail (or more likely, the drain). I have no circuit in mind...it's just a thought at the minute.  CM |
|
|
|
Post by ChrisK on Oct 9, 2006 20:15:57 GMT -5
It ain't the best, but it's easy. I would agree if the "input cap" is a local filter for the supply rail. If it's the actual signal input to the gate, no cap is needed since both of the resistors appear in parallel for AC signal analysis purposes. There are many. The VN2222 (wow, just like the 2N2222) is just one. These are much more common than depletion mode devices. They often have a body diode and are more often used for power (load) switching. Go to DigiKey and use their search engine. rocky.digikey.com/WebLib/On-Semi/Web%20Data/VN2222LL.pdfIt would be just like the circuit with two input resistors, one to common and one to the supply. No source resistor is needed unless you want to do a source follower (gain = +1). Well, I don't remember what "enhancement mode" actually means. Here's a data sheet www.onsemi.com/pub/Collateral/NTMD6N02R2-D.PDFThese tend to have a Vgs threshold of +1.5 VDC or so. |
|
|
|
Post by JohnH on Oct 9, 2006 22:02:55 GMT -5
It ain't the best, but it's easy. I would agree if the "input cap" is a local filter for the supply rail. If it's the actual signal input to the gate, no cap is needed since both of the resistors appear in parallel for AC signal analysis purposes. Chris - what I was getting at was, the cable buffer gets plugged into a guitar, which provides a variable resistance to ground of anything from zero to several hundred k. If the JFET gate is biased to ground only, then this is OK, but if it needs biasing to a different voltage, then the guitar resistance will mess with the bias, unless a decoupling capacitor on the input was provided, as with the built-in circuit. cheers John |
|
|
|
Post by ChrisK on Oct 10, 2006 13:57:19 GMT -5
Correct you are, sir! But, with the resistors being large (10 mOhm), the cap can be small.  |
|
|
|
Post by UnklMickey on Oct 10, 2006 20:10:35 GMT -5
...My requested for an 'enhancement mode N Channel FET still
stands though. I think it would be useful in battery circuits
because the gate is more positive than the source and we
wouldn't need to 'waste' supply voltage 'dropping' the
required negative Vgs across a source resitor.
We should finish up with more 'headroom' depending on the design.... this is either very useful, or totally irrelevant depending on application. in the case of a unity gain buffer, it's irrelevant. headroom won't be an issue, unless you are working with a device that is way off the mean. in a configuration with gain, an enhancement mode device, with the drain connected directly to ground would be very desirable. proper biasing would lead to the ability to reach much higher output, than a design using a depletion mode device, with the requisite drain resistor. additionally, without the degeneration caused by the drain resistor, higher gain would be feasible in a single stage. (of course, we could have mitigated the degeneration caused by the drain resistor by bypassing it.) on the other hand, a depletion mode device could have it's drain connected directly to ground. but, that would require a second supply, to bias the gate negatively. probably not a great choice, only worth mentioning, because it can be done. unk |
|
Channelman
Meter Reader 1st Class
Posts: 62
Likes: 0
|
Post by Channelman on Oct 11, 2006 6:58:34 GMT -5
Yep unk, you are correct all the way, as usual. And thanks Chris for the links for a VN2222. I didn't realise we have a Digikey over here....but we do at www.uk.digikey.com/I've been 'chewing an idea over' of a fet buffer with gain using an enhancement mode fet but apart from the fet, the biggest stumbling block to gain is the 9v (or even less) battery supply voltage With this supply we obviously can't have an output greater than 4.5 volts peak, and with a possible input of 1 volt peak, the gain must be limited to x4.5 This would be with everything ideal. In practice, after allowing for device spread, tolerances and low battery voltage this would be an overall in-out gain of about x3 Still 'kicking ideas over'  (Hick) CM |
|
|
|
Post by JohnH on Oct 11, 2006 7:39:28 GMT -5
Theres no fundamental reason to be limited to the generic 9V. Theres a tiny 12V battery that is used for garage door openers. Or, two 9V batteries = 18V Also, if one is providing that much boost, on a full signal, it is probably for the purpose of overdriving the amp. If the JFET clips it can contribute, and can be a good sound, very similar to tube clipping depending how it is set up. Thats actually my main interest in JFETs, searching for 'the sound'
John
|
|
|
|
Post by ccoleman on Oct 12, 2006 14:12:17 GMT -5
|
|
|
|
Post by JohnH on Oct 13, 2006 14:13:57 GMT -5
That Shaka circuit is not subtle! Those three JFET stages end to end add up to huge gain, and probably pretty much square waves on the output. What I'm going for is a wider range. Here is my current prototype and pet project:  It is also a multi-stage JFET circuit, with careful attention to where the gain is provided, the current and biasing through each stage, adjustable pre and post gain EQ, adjustable diode clipping and an unsusual tone stack. I don't normally make claims that I am not willing to substantiate. Until now... With low overdrive, it can go from clean boost with tone eq, through clean with a bluesy edge, up to a vintage plexi-crunch. Hitting the OD switch engages Angus sounds, and with partial clipping and max gain, boosts it to bring forth GNR 'slash' solo and 'scooped metal' possibilities. I'm still searching John |
|
|
|
Post by ccoleman on Oct 14, 2006 11:21:59 GMT -5
that combined with a 4real sustainer seems like it would make an absolutely deadly good rig !
what else are you searching for ?
|
|
|
|
Post by JohnH on Oct 14, 2006 18:30:31 GMT -5
Well there is always extra tweaking that is necessary because it is possible. And with no deadlines set by marketing departments and production directors, it becomes imperative. There is one major omission from my box - an LED indicator! Perhaps not sonically interesting, but practically very important. Adding it is possible but not trivial, unless I order some 3-pole stomp switches instead of the 2-pole ones I can buy locally. Going back to buffers, I found this yesterday: www.guitar-repairs.co.uk/jfet_guitar_preamp.htmThis preamp is available as a kit for rather alot of money. Generously, the developer has also posted the schematic. www.guitar-repairs.co.uk/FETPreamp.pdfit looks relatively high current, and I'm curious about the merits of the second bipolar transistor. This is described as being to lower output impedance. But the JFET is working already at quite a high current, so should be fairly low impedance. The bipolar is not an independent buffer stage, but is intimately tied in with the JFET, acting in a way that is not obvious to me, but may well be very clever!. Also, check out the diode across the supply cap. Reverse voltage protection by shorting out the power rail? curious... John |
|
Channelman
Meter Reader 1st Class
Posts: 62
Likes: 0
|
Post by Channelman on Oct 15, 2006 5:17:42 GMT -5
John, it's a dc pair whose voltage gain is defined by (R7 + R6)/R6, which gives it a calculated gain of x1.83 or 5.26dB. In practice with a 'middle of the road' J201 it turns out at x1.6 or 4.1dB. Like you said, high current.... the FET is taking 0.6mA and the Bipolar 1.9mA (Total 2.5mA). It will be low output impedance (well under 1 kohm due to negative feedback). This arrangement is a neat way of stopping the anti-phase signals on the drain and source restricting the 'headroom' since all the voltage gain is in the bipolar transistor, the drain hardly moves. This configuration warrants further thought. I'm sure it can be done better, at a lower current and definitely at a lower price. >John said...Reverse voltage protection by shorting out the power rail?Shorted via a 100 ohm...better a flat battery than duff transistors...eh? And a diode in series 'throws' 0.7 volts away from the rail. EDIT I've just had a look at the installation instructions and the guy says to expect 1000 hrs battery life but with an Energizer Battery at 625mAH I think that 250 hours would be nearer at 2.5mA drain. Also, and this is really strange. He fits it between the PU Switch and the Volume Pot. This is OK for not loading the PUs but no good for driving the cable....not from a 500k pot. If he changed the Vol Pot to 50k, OK...but no change, unless I've missed something. Well that's how it is on 'Strat type' with One Volume Control. On Gibson type with two Vol Pots it drives the cable, which I think is correct. Have a look, here's the link. www.guitar-repairs.co.uk/pdf/jfet_guitar_preamp.pdfCM |
|
|
|
Post by JohnH on Oct 15, 2006 15:58:04 GMT -5
I agree, the Strat installation loses the low impedance output benefit once the volume is turned down. It would be just as easy to put the amp after the volume pot.
Also, on batteries, 625mAH gets it down to 4.8V, by which time it may not be a happy circuit. The overall configuation is interesting though - worthy of exploration.
John
|
|
|
|
Post by ccoleman on May 27, 2007 6:00:01 GMT -5
Before building our JFET buffer I need to ask this question to the group... I heard about the Fender reissue Andy Summers (the Police) 61 tele modified, as Andy played it during the Police years, with a built-in overdrive, and wonder how easy it would be to modify our buffer circuit to give a clean boost that would overdrive the input of the next FX or Amp in the chain... Is a clean boost the most useful ?! Versus... would we rather want the boosted signal to actually overdrive the JFET and make a nice distortion before even exiting the guitar ?!
|
|
|
|
Post by JohnH on May 28, 2007 4:41:46 GMT -5
For a simple buffer with gain, I really like this Fetzer valve circuit: www.runoffgroove.com/fetzervalve.htmlThe designers have worked out all the parameters to make it match the response of a tube input stage, based on measuring the individual JFET characteristcs. I tried one with a J201, and have it built into one of my guitars. It runs at about 120 uA, but can make a significant boost to push the amp. I havnt got it to overdrive itself however, it stays clean, even with my three in series combo. I've also tried two or three stages to give in-guitar overdrive, but it got too fiddly so I moved to putting them in boxes like normal pedals (see above). So inside a guitar, I'm in favour of 'clean boost' (if wanted), or 'no boost' buffers . John |
|
|
|
Post by JFrankParnell on Jun 17, 2010 21:02:40 GMT -5
Ok, I dont know a JFET from Mos Def. In the cable preamp, which is the guitar end? On the built in amp, could you please post a wiring diagram? I would also like to see what it would look like with a toggle instead of the jack-switch. power it with photovoltaic cells on the front of your guitar. [...] the only problem is, if there's a power failure, the lights will go out, and you'll have no sound! .........................................................hey, wait a minute! You sound like the guy who appreciated the moon more than the sun because the moon gives light at night when you really need it but the sun gives light in the day when its already light out ;D |
|
|
|
Post by sumgai on Jun 17, 2010 22:06:44 GMT -5
.... power it with photovoltaic cells on the front of your guitar.
[...]
the only problem is, if there's a power failure, the lights will go out, and you'll have no sound!
.........................................................hey, wait a minute!
You sound like the guy who appreciated the moon more than the sun because the moon gives light at night when you really need it but the sun gives light in the day when its already light out ;DYes, he did have a sense of humor that kept us scratching our heads. Not to mention that he knew what he was talking about, and it was a rare day indeed when ChrisK, John or I could catch him out in an error. Unfortunately, he "disappeared" from these environs in September of 2007 - no notice, no forwarding address, nuttin'. The real Nutz here will tell you that we miss him, pure and simple. |
|
|
|
Post by JohnH on Jun 18, 2010 23:51:51 GMT -5
|
|
|
|
Post by ashcatlt on Jun 19, 2010 0:24:16 GMT -5
...the jack goes into the guitar. This jack ... Doncha mean the plug there, Jack...I mean John...I mean... The part that fits into the guitar goes there. The male end. The plug. |
|
|
|
Post by JFrankParnell on Jun 19, 2010 2:14:38 GMT -5
...the jack goes into the guitar. This jack ... Doncha mean the plug there, Jack...I mean John...I mean... The part that fits into the guitar goes there. The male end. The plug. lol, Yes, the plug is the male, and "she's got the jack" |
|
|
|
Post by ashcatlt on Jun 19, 2010 15:51:18 GMT -5
I don't think that's what Bon was talking about there, but that is a great way to remember it.
|
|
